1. Technical Field
This invention relates to a method for deciding a viscosity in case of correcting a density of a fluid test sample in accordance with the viscosity in the density measurement by means of an oscillating densimeter.
2. Description of the Prior Art
An U-shaped tube which is filled at the base ends (the both ends of the upper portion of U) is filled with a test sample (liquid or gas). When the mechanical vibration is given to the U-shaped tube, the tube vibrates in response thereto at the frequency based on the density of the test sample filled inside of the U-shaped tube. Therefore, the viscosity of the test sample can be obtained by measuring the frequency (or the oscillation period). It is the oscillating densimeter that the method of this instant application is applied to.
FIG. 2 shows a basic constitution of a conventional oscillating densimeter. A permanent magnet 11 is fixed on the pointed end (the base portion) of the U-shaped tube 10. Then the electric current with a specific frequency flows into a driving coil 14 provided nearby the permanent magnet the U-shaped tube 10 vibrates.
The vibration of the U-shaped tube 10 is detected by a sensor 13. The detected signals from the sensor 13 is amplified by an amplifier 12 and sent back to the driving coil 14. According to the above constitution, the U-shaped tube is to resonate with the frequency of the electric current in the driving coil. And the oscillating period of the U-shaped tube is measured in accordance with the output signals from the sensor 13, the results of which is offered to the density operation by the calculation means 15.
The density which is obtained easily from the oscillating frequency as described above has the difference based on the viscosity of the test sample as shown in FIG. 4. Namely, the larger the viscosity xcex7 is, the larger the difference rate represented by xcex94xcfx81/xcfx811 becomes (xcex94xcfx81=xcfx811xe2x88x92xcfx810, where the true value of the density is represented by xcfx810 while the measured value is represented by xcfx811).
In order to correct the density difference based on the viscosity xcex7, it is necessary to detect the viscosity of the test sample. The viscosity xcex7 can be detected in accordance with the relation between the viscosity and the attenuation constant, which is disclosed in IEE TRANSACTIONAL ON INDUSTRIAL ELECTRONICS AND CONTROL INSTRUMENTATION, VOL. IEC1-27, NO. 3, AUGUST 1980, 247-253 (the literature 1). Therefore, the density-attenuation constant characteristic at the 0 order oscillation can be represented by the function b0=f(xcex7) or that at the first order oscillation by the function b1=g(xcex7) respectively (the xe2x80x9corderxe2x80x9d will be explained later) as shown in FIG. 5. The 0 order oscillation is a mode that one node of the oscillation is positioned at the base ends of the U-shaped tube, which is shown as i=1 in FIG. 6. In general the density can be obtained in accordance with the frequency of this mode. And the first order oscillation is a mode that there are two nodes which are positioned at the base ends and the position at one quarter to the base ends, and is shovel as i=2 in FIG. 6. And there are also modes at higher order oscillation (i=3 and i=4, for example).
The conventional constitution is arranged in the literature 1 that the circuit as shown in FIG. 7 works in order to generate the oscillation at each mode.
Accordingly, the output from a piezoelectric element 21 as a sensor, that is to say a detected voltage Ud is inputted as signals U into a modulator 24 and a phase shift unit 25 via a variable gain amplifier 22 and a voltage-controlled phase adjuster 23.
And the signals U compose control signals Uc of the variable gain amplifier 22 via a rectifier 22 and an integrator 27. This loop, that is the variable gain amplifier 22xe2x86x92the voltage-controlled phase adjuster 23xe2x86x92the rectifier 26xe2x86x92the integrator 27xe2x86x92the variable gain amplifier 22, is provided with a function for fixing the height of the output U without regard to the height of the detected voltage.
The modulator 24 outputs a value which is an amplitude of the basic signals U multiplied by modulation coefficient xcex5, while the phase shift unit 25 outputs signals that has a phase of the basic signals U shifted to xe2x88x92xcex8 (45xc2x0, for example). Those two signals are added each other at a mixer 28, and then signals Uc can be obtained the phase of which is shifted for the angle corresponding to one of the modulation coefficient xcex510, xcex52 as shown in FIGS. 8a and 8b. By using the signals, the excited current Iexc is to be detected. Namely the signals U is made to delay by 45xc2x0 via the phase shift unit 25. On the other hands, the signals xcex51U or xcex52U (shown as xcex51 greater than xcex52, in FIG. 8a and 8b is multiplied by the modulation coefficient xcex51 or xcex52 respectively, and the product is added to the basic signals, in a result the signals Ue1 or Ue2 generates. In the signals Ue1 or Ue2, in case of xcex51 greater than xcex52, the delay angle is represented by xcex81 greater than xcex82.
And the modulation coefficient xcex5 can be changed by adjusting the value of N of the control signal xcfx89/N that is inputted into the modulator 24.
If the signals with sifted phases as above are obtained, a resonating frequency with new, phase can be obtained, too. And it is also possible to obtain the higher harmonics of the first order, that is the oscillation of i=2, or the oscillation of higher order.
When the relation between the viscosity and the attenuation constant is actually measured at the 0 order oscillation by the above constituted apparatus, the result is shown in FIG. 3(a). On the other hand, when it is measured at the first order oscillation, the result is shown in FIG. 3(b). FIG. 5 shows the graph combined the above two graphs together.
In order to correct the density difference based on the viscosity, if the viscosity-attenuation constant characteristic at the 0 order oscillation is represented by b0=f(xcex7), the attenuation constant has a peak at the point of the viscosity xcex71 (which is about 100 mPas in fact). This means that there are two values of the viscosity for the same attenuation constant. Therefore, it causes a trouble, that is which value should be used. If the viscosity-attenuation constant characteristic at the first order oscillation is represented by b1=g(xcex7) the viscosity value can be obtained unconditionally from the attenuation constant within comparative wide range. But since the attenuation constant gets peak at the point of the viscosity xcex7(which is about 700 mPas), it is confronted by the same problem as in case of using the 0 order oscillation.
Regarding the test sample, in case it can be predicted whether its viscosity is over or under 700 mPas, it is possible to make use of the viscosity obtained from the first order oscillation. But if the method for correcting the density difference is performed in accordance with the viscosity obtained from the first order oscillation although the viscosity at the 0 order oscillation was measured, there is a defect that it is impossible to expect the high accuracy.
This invention is proposed in consideration of the above conventional problems. The object of the present invention is to provide the method for deciding the viscosity at the specific order oscillation, even if there is a peak in the viscosity-attenuation constant characteristic at the specific order oscillation, by means of the viscosity-attenuation constant characteristic at the other order oscillation.
In order to achieve the above object, the present invention adopts the following method. Where there is a peak point of the attenuation constant in the viscosity-attenuation constant characteristic of the test sample at the specific order oscillation, the method is to decide whether the attenuation constant of the test sample at the other specific order oscillation is larger or smaller than the value of corresponding to said peak point, to decide according to said decision whether the viscosity detected by the attenuation constant obtained at the specific order oscillation belongs to either side lager or smaller than said peak point, and then to decide the viscosity of the test sample in accordance with the above decisions.
The 0 order which is applied directly to the densimeter is used as the specific order and the the first order is used as the other specific order. According to such arrangement, the viscosity measurement can be performed at the oscillation mode used by the density measurement. Therefore it is possible to expect the improvement of the accuracy of the viscosity measurement.